Skip to content

Latest commit

 

History

History
64 lines (35 loc) · 3.3 KB

havoc-OST.md

File metadata and controls

64 lines (35 loc) · 3.3 KB

Intro

This blog provides a intriduction to reviewing github repoitories of offensive security tools (OST) to help build comamnd and control (C2) framework detection rules for Shodan.

HTTP related artifacts that could be used for detection:

We searched for the "content-type" keyword to attempt to identify HTTP headers or rpofiles that could be used for detection.

SMB profile

First we identified an SMB profile for the Havoc framework

image

Possible rule: Content-Type: application/json; charset=utf-8 Server: Microsoft-HTTPAPI/2.0 X-Content-Type-Options: nosniff x-ms-environment: North Europe-prod-3,_cnsVMSS-6_26 x-ms-latency: 40018.2038 Access-Control-Allow-Origin: https://teams.microsoft.com

Havoc default HTTP header information

Within the GitHub repository we identified a number of hard coded values that can be used for detection. In particular the "X-Havoc" keyword in the header.

image

Possible rules:

"X-Havoc" https://www.shodan.io/search?query=%22X-Havoc%22 image

Expanding on the keyword we can make a more specific detection using the full HTTP header or the HTTP header hash.

To do this we can check the facet analysis section and understand how many unique headers make up the data set:

image

Using these headers we can now track specific Havoc header configs seperately which can in some cases allow us to track specific versions of the framework or in some cases specific activity groups or threat actors.

image

The new query leveraging the unique header hashes, would look like the below image:

image

Certificate based detection

In addition to headers we can also look for default certificate configuration to facilitate detetcion of the framework online.

image

Unfortunately, when looking at the facet analysis in Shodan there is limited uniqueness in some values to be able to use for a standalone detection.

Subject Name:

image

Issuer

image

JARM

The JARM however does allow us some detection opportunities:

image

"X-Havoc" ssl.jarm:"3fd21b20d00000021c43d21b21b43de0a012c76cf078b8d06f4620c2286f5e","3fd21b20d00000021c43d21b21b43d76e1f79b8645e08ae7fa8f07eb5e4202","2ad2ad16d2ad2ad0002ad2ad2ad2ad13962a56ecbfc3caaf51829946ab7fbe","40d1db40d0000001dc43d1db1db43d76e1f79b8645e08ae7fa8f07eb5e4202","40d1db40d0000001dc43d1db1db43de0a012c76cf078b8d06f4620c2286f5e","07d19d1ad21d21d00042d43d00000076e5b3c488a88e5790970b78ffb8afc2"